Day4

Uroš Godnov

Graphics

base: old system
lattice: creating function to plot
ggplot2

Base

example of a base graphics

index<-data.frame(year=2007:2016, pop=sample(10000:20000, size=10))

plot(index$year,index$pop)

Base - lines

type lines

#lines
plot(index$year,index$pop, type="l")

Base - histrogram

type histogram

plot(index$year,index$pop, type="h")

Base - title and axis labels

adding title and axis labels

plot(index$year,index$pop, type="l", main="Population by year",
     xlab="year",ylab="population")

Lab

use mtcars dataset and create a line plot where cyl must be on x axis and mean mpg on y axis

Lattice - creating a function

pop~year or mean mpg~cyl

lattice::dotplot(year~pop,data=index,main="Population by year",
        ylab="year",xlab="population")

Lattice - problem with y-values

we must change year values to factors

index$year<-as.factor(index$year)
lattice::dotplot(year~pop,data=index,main="Population by year",
        ylab="year",xlab="population")

Lattice - histrogram

histogram(year~pop,data=index)

Lattice - more complex formula

xyplot(mpg~wt | factor(cyl), data=mtcars, pch=19,
                main="MPG vs Wt", xlab="Wt/1,000",  ylab="MPG")

Lab

use mtcars dataset and create a barchart of mean mpg for number of gears per cyl

ggplot2

Foundation for many graphics applications:
- ggplot2
- Tableau
- Vega-Lite

ggplot2 - The idea

ggplot2

ggplot2

data: data.frame
aes: mapping data to properties; color, format and size
stats: transforms input variables to display values
scales: translate between variable ranges and properties ranges - categories -> colours; numbers -> positions
geoms: geometric objects (points, lines, …)
facets: forms a matrix of panels defined by row and column faceting variable
coordinates: defines the phiysical mapping of the aesthetics to the paper
themes: every part of the graphic that is not linked to the data

ggplot2 - basic object

which data to plot
which columns to use for x and y
how to draw the plot
“+” is used to combine the ggplot2 elements

ggplot(data=faithful,
       mapping=aes(x=eruptions,
                   y=waiting))+
  geom_point()

ggplot2 - different syntax

ggplot(data=faithful)+
  geom_point(mapping=aes(x=eruptions,
                   y=waiting))

ggplot()+
  geom_point(data=faithful,mapping=aes(x=eruptions,
                   y=waiting))

ggplot2 - adding colour

we can create subgroups in ggplot2 with colour parameter

ggplot(data=faithful)+
  geom_point(mapping=aes(x=eruptions,y=waiting,
                         colour=eruptions<3))

ggplot2 - different geometry

some geoms only need a single mapping and will calculate the rest for you

ggplot(data=faithful)+
  geom_histogram(mapping=aes(x=eruptions))

ggplot2 - many layers

layers are stacked int the order of code appearance

ggplot(data=faithful,
       mapping=aes(x=eruptions,
                   y=waiting))+
  geom_density_2d()+
  geom_point()

Lab

open ggplot2.txt. Solve excercises 1-4!

ggplot2 - statistics - 1

geom_bar uses stat_count() by default

ggplot(mpg) + 
  geom_bar(aes(x = class))

ggplot2 - statistics - 2

you can pre-calculate
and use stat = ‘identity’

mpg_counted <- mpg %>% 
  count(class, name = 'count')
  
ggplot(mpg_counted) + 
  geom_bar(aes(x = class, 
              y = count), 
        stat = 'identity')

ggplot2 - statistics - 3

using after_stat (ggplot2 ver3.3.0)
modifying mapping from stats

ggplot(mpg) + 
  geom_bar(aes(x = class, 
  y = after_stat(100 * count / 
                  sum(count))))

Values calculated by the stat is available with the after_stat() function inside aes()

Lab

open ggplot2 and solve the 5th excercise
additional hint:…+ stat_summary(aes(x = class, y = hwy),fun=?, geom = “point”, color=?)

ggplot2 - scales - 1

ggplot(mpg) + 
  geom_point(
  aes(x = displ, 
      y = hwy, 
  colour = class))

scales define how the mapping you specify inside aes() should happen. All mappings have an associated scale even if not specified
based on the vector type of a class,a discrete colour is picked

ggplot2 - scales - 2

we can take control by adding one explicitly
RColorBrewer::display.brewer.all()

ggplot(mpg) + 
  geom_point(aes(x = displ, y = hwy, colour = class)) + 
  scale_colour_brewer(type = 'qual')

ggplot2 - scales - 3

positional mappings (x and y) also have associated scales.

ggplot(mpg) + 
  geom_point(aes(x = displ, y = hwy)) + 
  scale_x_continuous(breaks = c(3, 5, 6)) + 
  scale_y_continuous(trans = 'log10')

Lab

open ggplot2 and solve the 6-7 excercises

ggplot2 - facet

the facet defines how data is split among panels. The default facet (facet_null()) puts all the data in a single panel
facet_wrap() and facet_grid() allows you to specify different types of small multiples
mind the scales

ggplot(mpg) + 
  geom_point(aes(x = displ, y = hwy)) + 
  facet_wrap(~ class)

Lab

open ggplot2 and solve the 8-9 excercises
scales and space can have free_x, free_y and free values

ggplot2 - coordinates - 1

the coordinate system is the fabric you draw your layers on in the end
the default `coord_cartesion provides the standard rectangular x-y coordinate system
changing the coordinate system can have dramatic effects

ggplot(mpg) + 
  geom_bar(aes(x = class)) + 
  coord_polar()

ggplot2 - coordinates - 2

changing mapping

ggplot(mpg) + 
  geom_bar(aes(x = class)) + 
  coord_polar(theta = 'y') + 
  expand_limits(y = 70)

ggplot2 - coordinates - 3

zooming with scales removes data outside limits

ggplot(mpg) + 
  geom_bar(aes(x = class)) + 
  scale_y_continuous(limits = c(0, 40))

ggplot2 - coordinates - 4

zooming with coords creates proper zoom

ggplot(mpg) + 
  geom_bar(aes(x = class)) + 
  coord_cartesian(ylim = c(0, 40))

ggplot2 - coordinates - 5

scale vs. coordinate transformation

scale<-ggplot(diamonds, aes(carat, price)) +
  geom_point() +
  scale_x_log10() +
  scale_y_log10()
coord<-ggplot(diamonds, aes(carat, price)) +
  geom_point() +
  coord_trans(x = "log10", y = "log10")
(scale | coord)

Lab

open ggplot2 and solve the 10th excercise

ggplot2 - themes - 1

theming defines the feel and look of your final visualisation
it is very easy to change looks with a prebuild theme
few themes in ggplot2 package
for more install ggthemes package

ggplot(mpg) + 
  geom_bar(aes(y = class)) + 
  facet_wrap(~year) + 
  theme_minimal()

ggplot2 - themes - 2

    ggplot(mpg) + 
      geom_bar(aes(y = class)) + 
      facet_wrap(~year) + 
      labs(title = "Number of car models per class",
           caption = "source: http://fueleconomy.gov",
           x = NULL,
           y = NULL) +
      scale_x_continuous(expand = c(0, NA)) + 
      theme_minimal() + 
      theme(
        text = element_text('Avenir Next Condensed'),
        strip.text = element_text(face = 'bold', hjust = 0),
        plot.caption = element_text(face = 'italic'),
        panel.grid.major = element_line('white', size = 0.5),
        panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank(),
        panel.ontop = TRUE
      )

ggplot2 - themes - 3

ggplot extensions

link

ggplot GUI

package esquisse
generates code
copy+paste code

ggplot2 - ggrepel - 1

the problem of labels, they usually overlap
geom_text_repel()
geom_label_repel()

ggplot(mtcars, aes(x=wt, y=mpg))+geom_point()+
  geom_label(label= row.names(mtcars))

ggplot2 - ggrepel - 2

library(ggrepel)

ggplot(mtcars, aes(x=wt, y=mpg, label= row.names(mtcars)))+
  geom_point()+
  geom_label_repel()

Lab

Take dataframe iris. Create scatter plot for Sepal.Length and Sepal.Width and label points with Species. Make sure the labels don’t overlap.

ggplot2 - ggforce - 1

adding missing functions, i.e. facets

library(ggforce)
ggplot(mtcars, aes(hp, mpg, colour = as.character(cyl))) +
  geom_point()+
  facet_zoom(x = cyl == 6)+theme(legend.title = element_blank())

ggplot2 - extensions

plot composition:
- gridExtra::grid.arange()
- ggpubr::ggarange()
- patchwork

ggplot2 - plot composition - 1

p1 <- ggplot(msleep) + 
  geom_boxplot(aes(x = sleep_total, y = vore, fill = vore))
p2 <- ggplot(msleep) + 
  geom_bar(aes(y = vore, fill = vore))
p3 <- ggplot(msleep) + 
  geom_point(aes(x = bodywt, y = sleep_total, colour = vore)) + 
  scale_x_log10()

ggplot2 - plot composition - 2

combining them with patchwork is a breeze using the different operators

(p1 | p2) / 
   p3

ggplot2 - plot composition - 3

combining them with patchwork is a breeze using the different operators
plot_layout(guides = ‘collect’) collects the legends

p_all <- (p1 | p2) / 
            p3
p_all + plot_layout(guides = 'collect')

Lab

Patchwork will assign the same amount of space to each plot by default, but this can be controlled with the widths and heights argument in plot_layout(). This can take a numeric vector giving their relative sizes (e.g. c(2, 1) will make the first plot twice as big as the second). Modify the code below so that the middle plot takes up half of the total space:

p <- ggplot(mtcars) + 
  geom_point(aes(x = disp, y = mpg))
p + p + p

ggplot2 - animation - 1

gganimate package
gganimate extends the API and grammar to describe animations
many different transitions that control how data is interpreted for animation

ggplot(economics) + 
  geom_line(aes(x = date, y = unemploy))

ggplot2 - animation - 2

transition_reveal: reveal data along a given dimension

ggplot(economics) + 
  geom_line(aes(x = date, y = unemploy)) + 
  transition_reveal(along = date)

ggplot2 - animation - 3

transition_states: transition between several distinct stages of the data

ggplot(mpg) + geom_bar(aes(x = factor(cyl))) + 
  labs(title = 'Number of cars in {closest_state} by number of cylinders') + 
  transition_states(states = year) + enter_grow() + 
  exit_fade()

Lab

In the animation below (as in all the other animations) the changes happens at constant speed. How values change during an animation is called easing and can be controlled using the ease_aes() function. Read the documentation for ease_aes() and experiment with different easings in the animation, i.e. ease_aes(“bounce-in-out”)

    mpg2 <- tidyr::pivot_longer(mpg, c(cty,hwy))
    ggplot(mpg2) + 
      geom_point(aes(x = displ, y = value)) + 
      ggtitle("{if (closest_state == 'cty') 'Efficiency in city' else 'Efficiency on highway'}") + 
      transition_states(name)